March 2004 - Inkscape Devel - Inkscape Mailman

RE: [Inkscape-devel] simplify (rect --> path) ???
by bulia byak 26 Mar '04

26 Mar '04

>I think simplify should keep the element as it is. An rectangle should >keep as a rectangle etc. This is not possible. A simplification is, by definition, a distortion. It may remove nodes and change their coordinates. On the other hand, a rect is, by definition, precisely a rect, mathematically correct without any distortions. You can't have it both ways :) Thanks for the patches, I will now review them. _________________________________________________________________ MSN Premium: Up to 11 personalized e-mail addresses and 2 months FREE* http://join.msn.com/?pgmarket=en-ca&page=byoa/prem&xAPID=1994&DI=1034&SU=ht…

1 0

more on win32 locale
by Adib Taraben 26 Mar '04

26 Mar '04

Hello, I found two other issues that addresses the win32 locale bug: 1. When drawing a polygon (the star) numbers in the points element are locale dependent, strange the star is drawn correctly. 2. The root document with and height are also locale dendent. I found the problem that causes #1 but I didn't found the source that writes the root width and height with the units. There are so many functions with printf that I must think some of them are never called? Adib. cvs diff sp-polygon.cpp (in directory C:\projekte\inkscape\src) Index: sp-polygon.cpp =================================================================== RCS file: /cvsroot/inkscape/inkscape/src/sp-polygon.cpp,v retrieving revision 1.2 diff -r1.2 sp-polygon.cpp 20a21 > #include "stringstream.h" 99c100 < GString *result; --- > Inkscape::SVGOStringStream os; 101d101 < char *res; 105,106d104 < result = g_string_sized_new (40); < 112c110 < g_string_sprintfa (result, "%g,%g ", bpath [i].x3, bpath [i].y3); --- > os << bpath [i].x3 << "," << bpath [i].y3 << " "; 120,123c118,119 < res = result->str; < g_string_free (result, FALSE); < < return res; --- > > return g_strdup (os.str().c_str()); ***** CVS exited normally with code 1 *****

1 0

simplify (rect --> path) ???
by Adib Taraben 26 Mar '04

26 Mar '04

Hello, during my work on simplify output I found that simplify destroys the basic structure of the element. The output is always a path. simplify on a rect-object becomes a path-object. circle will be path. I think simplify should keep the element as it is. An rectangle should keep as a rectangle etc. Adib.

1 0

win32 locale bugs patches that work
by Adib Taraben 26 Mar '04

26 Mar '04

Hello, I have some patches for the win32 locale bugs (floating point output with komma instead point) I made the diffs diff -u4 oldfile nexfile > difffile I am using a german winXP. splivarot addresses simplify path repr-util addresses rect, circle output x and y etc. coordinates with komma instead point style addresses rect, cirlce output opacity etc. with komma instead point I don't have access to the bugs page, so I don't know the numbers yet. HTH, Adib. --- style.cpp 2004-03-26 16:16:46.000000000 +0100 +++ style.adib.cpp 2004-03-26 16:10:56.000000000 +0100 @@ -36,8 +36,9 @@ #include "marker-status.h" #include "uri-references.h" #include "sp-paint-server.h" #include "style.h" +#include "stringstream.h" namespace Inkscape { @@ -2044,14 +2045,17 @@ */ static gint sp_style_write_ifloat (gchar *p, gint len, const gchar *key, SPIFloat *val, SPIFloat *base, guint flags) { + Inkscape::SVGOStringStream os; + if (((flags & SP_STYLE_FLAG_IFSET) && val->set) || ((flags & SP_STYLE_FLAG_IFDIFF) && val->set && (val->value != base->value))) { if (val->inherit) { return g_snprintf (p, len, "%s:inherit;", key); } else { - return g_snprintf (p, len, "%s:%g;", key, val->value); + os << key << ":" << val->value << ";"; + return g_strlcpy (p, os.str().c_str(), len); } } return 0; } @@ -2062,15 +2066,17 @@ */ static gint sp_style_write_iscale24 (gchar *p, gint len, const gchar *key, SPIScale24 *val, SPIScale24 *base, guint flags) { + Inkscape::SVGOStringStream os; if (((flags & SP_STYLE_FLAG_IFSET) && val->set) || ((flags & SP_STYLE_FLAG_IFDIFF) && val->set && (val->value != base->value))) { if (val->inherit) { return g_snprintf (p, len, "%s:inherit;", key); } else { - return g_snprintf (p, len, "%s:%g;", key, SP_SCALE24_TO_FLOAT (val->value)); + os << key << ":" << SP_SCALE24_TO_FLOAT(val->value) << ";"; + return g_strlcpy (p, os.str().c_str(), len); } } return 0; } --- splivarot.cpp 2004-03-26 16:16:04.000000000 +0100 +++ splivarot.adib.cpp 2004-03-26 14:58:33.000000000 +0100 @@ -17,8 +17,9 @@ #include <libart_lgpl/art_misc.h> #include "xml/repr.h" #include "svg/svg.h" +#include "stringstream.h" #include "sp-path.h" #include "sp-text.h" #include "style.h" #include "inkscape.h" @@ -1544,24 +1545,23 @@ gchar * liv_svg_dump_path (Path * path) { - GString *result; - result = g_string_sized_new (40); + Inkscape::SVGOStringStream os; for (int i = 0; i < path->descr_nb; i++) { Path::path_descr theD = path->descr_cmd[i]; int typ = theD.flags & descr_type_mask; if (typ == descr_moveto) { Path::path_descr_moveto* nData=(Path::path_descr_moveto*)(path->descr_data+theD.dStart); - g_string_sprintfa (result, "M %lf %lf ", nData->p[0], nData->p[1]); + os << "M " << nData->p[0] << " " << nData->p[1] << " "; } else if (typ == descr_lineto) { Path::path_descr_lineto* nData=(Path::path_descr_lineto*)(path->descr_data+theD.dStart); - g_string_sprintfa (result, "L %lf %lf ", nData->p[0], nData->p[1]); + os << "L " << nData->p[0] << " " << nData->p[1] << " "; } else if (typ == descr_cubicto) { Path::path_descr_cubicto* nData=(Path::path_descr_cubicto*)(path->descr_data+theD.dStart); @@ -1570,32 +1570,34 @@ NR::Point tmp = path->PrevPoint (i - 1); lastX=tmp[0]; lastY=tmp[1]; } - g_string_sprintfa (result, "C %lf %lf %lf %lf %lf %lf ", - lastX + nData->stD[0] / 3, - lastY + nData->stD[1] / 3, - nData->p[0] - nData->enD[0] / 3, - nData->p[1] - nData->enD[1] / 3, nData->p[0],nData->p[1]); - } + os << "C " << lastX + nData->stD[0] / 3 << " " + << lastY + nData->stD[1] / 3 << " " + << nData->p[0] - nData->enD[0] / 3 << " " + << nData->p[1] - nData->enD[1] / 3 << " " + << nData->p[0] << " " + << nData->p[1] << " "; + } else if (typ == descr_arcto) { Path::path_descr_arcto* nData=(Path::path_descr_arcto*)(path->descr_data+theD.dStart); - // g_string_sprintfa (result, "L %lf %lf ",theD.d.a.x,theD.d.a.y); - g_string_sprintfa (result, "A %g %g %g %i %i %g %g ", nData->rx,nData->ry, nData->angle, - (nData->large) ? 1 : 0,(nData->clockwise) ? 0 : 1, nData->p[NR::X],nData->p[NR::Y]); + os << "A " << nData->rx << " " + << nData->ry << " " + << nData->angle << " " + << ((nData->large) ? "1" : "0") << " " + << ((nData->clockwise) ? "0" : "1") << " " + << nData->p[NR::X] << " " + << nData->p[NR::Y] << " "; } else if (typ == descr_close) { - g_string_sprintfa (result, "z "); + os << "z "; } else { } } - char *res; - res = result->str; - g_string_free (result, FALSE); - - return res; + return g_strdup (os.str().c_str()); + } --- repr-util.cpp 2004-03-26 16:24:02.000000000 +0100 +++ repr-util.adib.cpp 2004-03-26 15:35:13.000000000 +0100 @@ -29,8 +29,9 @@ #include <glib.h> #include "repr-private.h" +#include "stringstream.h" /*##################### # DEFINITIONS #####################*/ @@ -630,21 +631,16 @@ unsigned int sp_repr_set_double (SPRepr *repr, const gchar *key, double val) { - gchar c[32]; - + Inkscape::SVGOStringStream os; + g_return_val_if_fail (repr != NULL, FALSE); g_return_val_if_fail (key != NULL, FALSE); - //## This can be done in several ways. g_ascii_formatd() - //## might be the best for now. g_snprintf() can be - //## affected by LOCALE. g_ascii_formatd() is always "C" locale. - //g_snprintf (c, 32, "%.8f", val); - //sp_xml_dtoa (c, val, 8, 0, FALSE); - g_ascii_formatd (c, sizeof (c), "%.8f", val); - - return sp_repr_set_attr (repr, key, c); + os << val; + + return sp_repr_set_attr (repr, key, os.str().c_str()); } unsigned int sp_repr_set_double_default (SPRepr *repr, const gchar *key, double val, double def, double e)

1 0

Updated spanish translation
by Lucas Vieites Fariña 26 Mar '04

26 Mar '04

Hi, I just uploaded a 'patch' with the spanish po file. As soon as I get a little extra free time I'll finish the tutorial. See you, -- Lucas Vieites Fariña Dpto. Desarrollo <lucas@...212...> Asix Informática http://www.asixinformatica.com/ Tel/Fax: +34 986 54 26 98

1 0

RE: [Inkscape-devel] Release
by bulia byak 25 Mar '04

25 Mar '04

Ted, first, there's a compilation errror in db.cpp: db.cpp:230: error: redefinition of `GSList* Inkscape::Extension::DB::get_output_list()' db.cpp:217: error: `GSList* Inkscape::Extension::DB::get_output_list()' previously defined here db.cpp:230: error: no `GSList* Inkscape::Extension::DB::get_output_list()' member function declared in class `Inkscape::Extension::DB' make[4]: *** [db.o] Error 1 make[4]: Leaving directory `/home/d/ink/inkscape/src/extension' make[3]: *** [all-recursive] Error 1 make[3]: Leaving directory `/home/d/ink/inkscape/src/extension' make[2]: *** [all-recursive] Error 1 make[2]: Leaving directory `/home/d/ink/inkscape/src' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/home/d/ink/inkscape' make: *** [all] Error 2 Now on bugs: > > >857316 - win32: non-english locales: font style matching broken > > >852976 - win32: mouse becoming unresponsive > >So we need a couple of replacements for these then. I noticed there are >a couple of Mac OS X bugs that were put in the repository - has anyone >had a chance to look at those? Besides there are some that are unconfirmed: https://sourceforge.net/tracker/index.php?func=detail&aid=911292&group_id=9… Unconfirmed, against 0.37, and seems to be a GTK issue anyway https://sourceforge.net/tracker/index.php?func=detail&aid=908565&group_id=9… No reproducible steps, against 0.37; the crash itself is 99% fixed by the new rasterizer, although maybe not the underlying cause (hard to tell) https://sourceforge.net/tracker/index.php?func=detail&aid=915445&group_id=9… Not reproducible, the original report used a broken font. Works for me. https://sourceforge.net/tracker/index.php?func=detail&aid=922904&group_id=9… Works with the latest build on my Windows 98. If anyonce could test some of these bugs to resolve them one way or the other that would be helpful. _________________________________________________________________ STOP MORE SPAM with the MSN Premium and get 2 months FREE* http://join.msn.com/?pgmarket=en-ca&page=byoa/prem&xAPID=1994&DI=1034&SU=ht…

1 0

RPM's now in fedora.us
by Peter Linnell 25 Mar '04

25 Mar '04

Hi all, Just a note to let you know, Inkscape 0.37 RPMS have been approved by Fedora QA and is now in the fedora.us repsitory under stable - aka Fedora Extra's. The current Fedora rpms on the sourceforge page are the same packages, just built by me. This will be more important with the release of Fedora Core2, as it will be already configured to download and install Inkscape and other packages in stable. Now, that there is a good spec file, updating to 0.38 should be trivial. Peter

1 0

RE: [Inkscape-devel] Release
by bulia byak 25 Mar '04

25 Mar '04

>Currently in SF I see that we have 33 bugs out of 252 filed, so we need >to fix eight of them. Ironically, we have 8 high level bugs - so I >think that fixing those should be our target for the release. Agreed, although I'm not sure we can fix all remaining high-level bugs soon. I think we'll have to leave a couple of them standing for the release (fixing the corresponding number of minor ones instead). >915374 - Command Line Printing doesnt work There's a Sodipodi patch in our tracker. Someone who has a printer and can test it should take on this one. >914611 - inkview crash on startup I haven't looked but this one should be easy. >866103 - non-well-formed preferences.xml causes data loss Bryce was going to look into that one. What needs to be done is simply to back up the prefs file before saving, so it should not be difficult. >857316 - win32: non-english locales: font style matching broken That's difficult and may not be worthwhile to fix before we switch the font system to Pango anyway. >857311 - win32: correct output SVG of numerics is locale-dependent I'm working on that. Should be ready for the release. >852976 - win32: mouse becoming unresponsive This one is difficult. I found no easy way to fix it. The problem is that as soon as grabbed item goes beyond the screen edge, it stops receiving mouse events (although it's still grabbed). The top-level window receives the events instead, but it's difficult to dispatch those events in the window code because for this it must know exactly what type of object is grabbed (object, knot, guideline) and what handler to call. Annoyingly, Sodipodi does not have this problem on Windows for some reason. We really must brainstorm this one. _________________________________________________________________ Check out MSN PC Safety & Security to help ensure your PC is protected and safe. http://specials.msn.com/msn/security.asp

2 1

Re: [Inkscape-devel] tutorial news
by Cédric GEMY 25 Mar '04

25 Mar '04

Thanks, I'll compare this to my translation and take in charge to pass it into french. SOON AVAILABLE. Cédric

1 0

RE: Gradient Colour Bug?
by bulia byak 24 Mar '04

24 Mar '04

>When I change the gradient colours in any one object, every object gradient >changes to the new colours. It happens for linear and radial gradients. This is as designed. When you assign the same gradient to more than one object, they share the gradient (but not position/direction). If you want a new object to use an independent copy of the gradient, press the Add button on the fill/stroke dialog. _________________________________________________________________ http://join.msn.com/?pgmarket=en-ca&page=byoa/prem&xAPID=1994&DI=1034&SU=ht…

2 1